---
metadata:
  - name: generator
    content: Diplodoc Platform v5.50.6
alternate:
  - https://ytsaurus.tech/docs/en/flow/yql/features.md
  - https://ytsaurus.tech/docs/ru/flow/yql/features.md
---
> **Documentation Index:** Fetch the complete configuration index at https://ytsaurus.tech/docs/en/llms.txt

<!-- source: en/_includes/flow/yql/features.md -->
# Supported YQL constructs in YTsaurus Flow

{% note info %}

This page describes YQL constructs specific to working with Flow. The general YQL syntax is described in the [YQL documentation](https://ytsaurus.tech/docs/en/yql/index.md).

{% endnote %}

## Row-wise stream transformation (map) {#map}

The main construct is `INSERT INTO ... SELECT ... FROM`. It reads a [stream](https://ytsaurus.tech/docs/en/flow/concepts/glossary.md#stream-and-computation) from an ordered dynamic table (a YTsaurus queue), applies a transformation, and writes the result to another queue.

Example of a query part:

```yql
INSERT INTO
    <cluster-name>.`//home/my-project/output_queues/sink_queue`
SELECT
    string_field || "_processed" AS processed_field,
    int64_field * 2 AS doubled,
    EndsWith(string_field, "bar") AS predicate
FROM
    <cluster-name>.`//home/my-project/input_queues/source_queue`
WHERE int64_field > 0;
```

You can work with files and UDFs (built-in and user-defined), lambda expressions, and code generation.

You can combine multiple `INSERT INTO ... SELECT` operations in a single query.


## Stream join with a dynamic table (lookup join) {#lookup-join}

You can join a stream with a sorted dynamic table (a key-value table). Supported join types for the “stream + KV table” pair are `LEFT`, `LEFT ONLY`, `LEFT SEMI`, and `INNER`. For the “KV table + stream” pair, the types are symmetric.

```yql
$input_stream =
    SELECT key, value, key || "_before" AS key_before
    FROM <cluster-name>.`//home/my-project/input_queues/source_queue`
    WHERE value > 2;

$joined_stream =
    SELECT
        left_arg.key AS key,
        left_arg.value AS value,
        left_arg.key_before,
        right_arg.kv_value
    FROM $input_stream AS left_arg
    INNER JOIN
        <cluster-name>.`//home/my-project/states/kv_table` AS right_arg
    USING (key);

INSERT INTO <cluster-name>.`//home/my-project/output_queues/sink_queue`
SELECT * from $joined_stream
WHERE value * 2 <= kv_value;
```

## Roadmap {#roadmap}

Under development:
- Aggregations over fixed-size windows (hopping windows)
- Join with static tables
- Join by prefix of key columns
- Join of multiple streams with each other

## See also

- [Computation (concept)](https://ytsaurus.tech/docs/en/flow/concepts/computation.md)
- [Quick start (YQL)](https://ytsaurus.tech/docs/en/flow/yql/getting-started.md)
<!-- endsource: en/_includes/flow/yql/features.md -->